5 replies to this topic

[gunji]

Hi everyone

 

I have a problem.

 I need to determine the upstream pressure of a restriction orifice. Unfortunately, it is a two phase flow and i don't know how to do it nor the company has the tool. I have stated that as we are 80% mass of liquid, using Liquid RO calculation is good but not perfect (I have a convergence). Unfortunately, the client thinks it is less conservative (and he is right) and prefers using gas calculation. Unfortunately, it does not converge. I have thought of 2-phase pressure relief valve calculation but the surface is very low.

 

I am out of option.

 

Any thoughts?

 

Thanks

[latexman]

Calculate average properties of the two phases and use those just like a single phase.

 

What is at risk if it's way off?

[katmar]

Many years ago I looked at using orifices as steam traps. At one stage orifices were favored over mechanical steam traps in applications where reliability was more important than efficiency (like warships).  I seem to remember that I got reasonable numbers applying the Lockhart-Martinelli procedure, but using standard orifice calculations in place of the standard pipe calculations usually used by L-M.  I guess this is similar to what latexman has suggested, just with a different method for weighting the liquid and vapor contributions.

[Bobby Strain]

A diligent search should yield the best procedure. Try whatever is suggested, including HEM method for relief valves. I recall some procedure from many years ago that calculated a size for vapor and a size for liquid and add them. And, what is the purpose of the orifice?If it is important for some reason, you should keep looking. Or hire a specialist.

 

Bobby

[breizh]

Hi,

take a look 

Breizh

[PaoloPemi]

as suggested by others probably you can obtain a quick estimate with a HEM (based) model (vapor and liquid at equilibrium, traveling at same speed),

in general, as for PSV, CV etc. there are many models which can include slip, non-equilibrium etc. but these are complex and may require empirical parameters difficult to estimate (see as example the paper attached to breizh's post), for these you can collect data (from measurements) and data regression...